Abstraction for Solving Large Incomplete-Information Games

ion for Solving Large Incomplete-Information Games Tuomas Sandholm Computer Science Department Carnegie Mellon University Most real-world games and many recreational games are games of incomplete information. Over the last ars, abstraction has emerged as a key enabler for solving large incomplete-information games. First, is abstracted to generate a smaller, abstract game that is strategically similar to the original game. Second, an approximate equilibrium is computed in ract game. Third, the strategy from the abstract game is mapped back to the original game. In this paper, I will review key developments in the I present reasons for abstracting games, and point issue of abstraction pathology. I then review the practical algorithms for information abstraction and action abstraction. I then cover recent theoretical breakthroughs that beget bounds on the quality of the strategy abstract game, when measured in the original game. I then discuss how to reverse map the opponent’s nto the...

• Type: Journal
• Year: 2015
• Where: AAAI
• Authors: Tuomas Sandholm
• Keywords: AAAI 2015, Intelligent Agents